First World Problems (an ongoing series)

Problem: The Hellscape Bird Site is bad for my mental health, the mental health of others, and probably society at large.

Problem: it’s where the “action” is, lots of people I know online are only there and nowhere else, and it is occasionally valuable for news esp. in the tech world.

I can’t leave and I can’t stay.

Solution one: unfollow lots of people. good but people are still going to retoot shit into your timeline.

Solution two: mute like a mofo. better maybe. manual effort.

Solution three: fix our culture. unlikely in the short term.

The Witcher is dumb and bad

Here is some interesting information about The Witcher.

This should have been a good show, probably. Maybe. Some other universe is currently like, “Game of what? This is the real deal.”

Not here, though. Not us.

I had some thoughts in the first few episodes.

For example, in episode one, I wrote:

What the fuck is happening. Who the fuck are these people. Is there a prequel I was supposed to have watched? Why does the nation of Nils Bohr hate Candyland and Discount Kirsten Dunst?

and it kinda just goes downhill from there.

The dialog is so out of place much of the time. I assume it’s not just weirdly translated from Polish but actually just written that way by native English speakers.

Wizard lady, whose origin story is she’s an ugly hunchbacked abomination, gets a gratuitous shot of her tits but it’s during the Special Magic Surgery to make her pretty (the actress is very pretty and has fantastic tits). But like, “Yeah, ok so, I fixed your hunchback and fucked up face, and the one leg that is longer than the other, and your fucked up arms, and twisted spine, and all the other stuff. Took a few minutes less than I’d planned, though, your tits are just, I mean, 11 outta 10. Didn’t need to do a thing. Just fucking great.”

We coulda seen her tits later, and just assumed they were not spectacular all along, you know?

And it just goes downhill from there.

There are moments I do enjoy, I guess. Occasionally Cavill does a decent job of being somewhat world-weary. But he only has one face so “world-weary seen-it-all Geralt” is exactly the same as “angry Geralt” which is exactly the same as “just nutted Geralt”. Dude has exactly one look. One-faced as hell.

Anyway sometimes the supporting characters are kinda fun, I guess. The scenery is OK. But there’s no cinematography of note in a show about this grand fantasy world. Compare with Mr. Robot, a goddamned triumph of using a camera for a TV show.

In the end, it’s like a fantastically high-budget “D&D actual play”, only with less love of the source material.

And better tits.

Pay for things dammit

Just a note, some things I pay for like a moron instead of expecting surveillance capitalism to give it to me for free:

Fastmail – Fuck Google, fuck Gmail.

Micro.blog – replacing WordPress as my blog platform when the subscription runs out.

Overcast – I don’t use a lot of the actual features, but it’s better than the Apple Podcast app, I guess.

Apollo – I don’t consider myself “a redditor” but I do browse reddit enough that having a real app is worth it.

Things I’ve experimented with paying for:

Evernote – it’s pretty OK I guess. I will pay for a year, and use it for 3-4 months then just go back to Notes on my phone, because my note-taking needs are so minimal. Then I’ll not use it for 3-4 months, feel shitty that I’m wasting my money, and use it. Then let the subscription expire. Then a year later, repeat. Currently, Notes sync doesn’t work, so I’m using it. Once (if?) Apple fixes Notes sync I’ll probably go back.

Remember The Milk – See “Evernote”, only for reminders and tasks. It’s a damn good app; my only complaint is they did a shit job on the iPad form factor. I just keep going back and forth with “my actual to-do list needs are SO. DAMN. MINIMAL. that I can get by with the shit Reminders app on my phone”. Or, alternatively, I only need an actually good to-do app in short bursts.

Hot Take: moving to Python 3 is fucking easy, you lazy shitbirds

https://engineering.khanacademy.org/posts/goliath.htm

Look, everyone’s codebase is different. There’s a shit ton of baggage, hidden complexity, 100% critical code with 0% test coverage, and so on and so on. I get it. I really do. But.

1. PEP-3000 was released in 2006. Python 3 was released in (late) 2008. You’ve had 10 years to plan your migration strategy.
2. It took a long time to get there, but there’s been a critical mass of important libraries ported to 3. Example 1, Example 2.

Here’s why porting to 3 is “really hard”, in most likely order:

Your devs are bored with python and want to pad their CVs with something new.
Your team has written shit code.
Your management won’t give you the time and space to do it.

That’s it. Some combination of those 3 things, most likely the first. You built a giant thing with Python and now it’s “done”, but it’s also boring. Or you just didn’t ever give a shit about designing your code for 3. Or maybe you did, you gave it a good try, but you just can’t get management to approve (new headcount/feature freeze/etc) you think you’ll need to do it right.

Switching platforms has as much to do with attracting (and retaining) talent, and getting your corporate blog post noticed in Orange Site, as it does things like “performance” and “time to market”.

Lots of codebases are fucked up and management is universally shitty but be real: you’re moving to (Go/Scala/Kotlin/Rust/whatever is hip this week) because you are bored with whatever you’re using now.

Flickr is going to die and it’s all your fault

The first nail in the coffin: https://techcrunch.com/2019/12/19/flickr-owner-smugmug-emails-subscribers-with-an-urgent-request-help-us-find-more-paying-users/

The tl;dr is, Flickr got bought away from Yahoo before they could kill it, and decided to do a strange thing: provide a service in exchange for money. There’s now a generally useless free tier, and a “pro” tier that costs money.

You know how this ends: they “cannot continue to operate it at a loss as we’ve been doing”.

You also know where I’m going with this: there’s 2 sides.

On the one hand, Flickr was already dead, killed by Yahoo! and their absolutely staggering incompetence. Yahoo! is so fucking dumb, a fucking poster-child for why capitalism is bullshit. They managed to make nothing but the most idiotic decisions over and over and over, and are only alive because they amassed a horde of fuck-you money. Yahoo! didn’t do a fucking thing with Flickr in response to Google Photos, Facebook/Instagram, the iPhone, etc.

A reminder for you kids: Flickr was Instagram but with an emphasis on actual photography, not face-tuned imbeciles. It was a complete (albeit somewhat simple) social network. Everything was there to compete with everything else. All it needed was leadership. It had none. There was no end but death.

The other side is that users just won’t accept paying money for a thing. Users have largely been trained to accept that uploading 100% of your sensitive, personal information for abuse by corporate forces is just the price you pay for photo-sharing. Time and time again users have stated they’d prefer free over for-pay, even if it means half of Moldova has their identity information.

That’s where I truly get mad. It is likely that Yahoos mismanagement killed Flickr before SmugMug even bought it. That said, SmugMug has done as close to 100% correct as you can hope for. They have made a legit effort to make Flickr what it was and what it could be. But the damage is done. “I paid for my cell phone and my cable connection, why should literally anything else cost me money”.

I hope some drunk asshole in a tracksuit empties your bank account and spends it on vodka and strippers.

Let’s all panic

God fucking dammit.

Yes all you need to turn an 80% AR lower into a gun is:

  1. An AR lower parts kit
  2. An upper with bolt carrier group and charging handle
  3. A stock or brace
  4. A pistol grip
  5. A machinists vice
  6. A drill press
  7. Assorted hand tools
  8. AR wrench
  9. Dremel tool

Oh and like, you’re probably going to have to try 3 or 4 times unless you’re an experienced gunsmith, armorer, or machinist.

Oh and after you’ve done all that, there’s no guarantee it’ll even work reliably. Depends on how good of a machinist you are.

I’ve got an 80% lower that is “in progress”. It’s in progress because

  1. I don’t have a lot of free time and
  2. ITS NOT FUCKING EASY

also it’s cheaper to just buy a fucking AR. Cost of a cheap AR carbine: $450. Cost of AR parts: I’m about $700 in, but you can go cheaper, if you try.

They’re not turning these out on the street. The real ones are cheaper and easier to obtain, will work reliably, and don’t require motherfucking power tools and literal hours of your life.

It’s too late to “fix” Twitter

Of course everyone by now has read about the whole “fix twitter” thing.

It is of course too late. You can read a good summary here. In short: Jack led the organization into these problems. The fear of the stock price dropping hemmed them in. They fell for bad-faith arguments from assholes (conservatives). No one wanted to lose their job over bold initiatives or worse, telling the boss he’s an asshole.

Now it’s something-something underpants-gnomes blockchain. It’s too late. The damage is done. No one can fix it. You made a cesspool with occasional good jokes. Time to move on.

For what it’s worth, I’m bearish on federation. You can look up all the reasons yourself – Orange Site has got you covered.

I was a backer of App.net back in the day. The idea was solid but it wasn’t the sort of thing that was built for a giant adware business. You (probably) couldn’t build a business with 5000 employees from the App.net model; you’d have a smaller set of (hopefully) sustainable businesses, or business units of a larger company.

Ultimately their “idea” was a portable social graph, and open APIs to build applications. Monetization was underpants-gnomeish and it failed. On the plus side it was genuinely an honest effort.

Dorsey’s effort reeks of Crypto Bro bullshit and his usual detached-from-reality schtick. I don’t trust it and I don’t have high hopes.

The ultimate fix is the internet: blogging, RSS, open protocols. Sadly that stuff doesn’t let a small group of investors get rich anymore, so we got rid of it.

A grab-bag of tweet responses

I don’t tweet any more, but I still hop into twitter from time to time to see what’s going on. It’s still a hellscape.

Taking shit on any kid is scumbag-level shit. That said, Greta is a “public persona” and Barron is not. He doesn’t give speeches, appear in media, whatever. The former is going to get shit-talked by virtue of being heavily in the public eye. The latter has seemingly chosen to not be in the public eye at all, so attacks on him are as loathsome as attacks on the Obama kids, Chelsea Clinton, etc.

HomonymGate was also still yet another proof that Republicans are shit-sucking scum.

Yeah no one cares.

 

What a world, what a world.

I will straight-up knock you off the escalator if you do this.

 

If it doesn’t run a browser, does it matter?

Consider Urbit. It’s a brand new … something.

[IMPORTANT: Please, when considering it, try and forget that its primary creator is a racist psychopath; and assume, for the moment, that none of his utterly bullshit ideas about “Dark Enlightenment” are somehow baked in.]

It’s full of some interesting ideas. I like the idea of a small, easily-understood dingus that does what you need it to, and only that, and that is at least in theory designed to last more or less forever.

Then there’s Snap. It’s Google trying to fix all the problems they have in datacenters at their scale. Lots of people have problems with datacenters and scale, though; the catch with Snap is it’s effectively getting rid of TCP/IP, by building a new networking protocol.

It is also interesting. Building things without baggage and with the complete knowledge of the past N years, combined with the resources to act on a “big” design, is a hell of a thing.

Both of these things really start from a clean slate, and present big ideas free of existing limits. They are, in their own way, an attempt to move to the Next Big Thing.

My question then is, since neither of them runs a web browser, who gives a fuck?

OK, OK, I know, a data center networking stack isn’t designed to run a browser. I think my point still stands though: the browser is still down in the muck of more-or-less 90’s technology. Yeah. it connects to big data and that’s where a lot of the “real work” happens, so it can be sort-of 90’s.

And yes, weird moonshots like Urbit are in a way just a side-effect of there being way too much money in Silicon Valley, such that it’s all just one big crapshoot, and we might as well fund more-or-less everything because no one knows fuck-all about what is actually going to work. Remember when Twitter seemed pointless and dumb?

[IMPORTANT: Twitter is pointless and dumb, but somehow we’re all stuck with it.]

Anyway, ordinary people will never run Urbit, and I’d wager that it won’t be able to hold onto the attention of anything but the most niche nerds for more than a bit. It doesn’t run the ubiquitous tool of computing: the web browser. Hell even their docs say they intend to get rid of the browser-based “Bridge” at some point. Naturally, I guess; they’re trying to imagine a baggage-free future, so why keep this one remnant of the past around.

My dad used to say that a computer wasn’t really a computer without a printer. This was back when you used computers to compute, then you printed the results on paper. Obviously before networks were commonplace (and wireless!). And now printers are just for legal documents, flyers, and D&D character sheets.

Same goes for things like USB sticks; I keep a few around for emergency safe/recovery boot. They’re sort of pointless for a lot of workaday problems now.

We let the baggage of the past go all the time. But we also don’t: every computer in my house runs some sort of Unix or Unix-like. We still have files as streams of bytes on a disk (although it’s not a spinner platter of rust any more). We still have graphics drivers and color depths and SMTP.

Does everything depend on a giant pile of yesterday’s baggage, which is why moonshots seem so weird and alien? Can we ever move past this stuff without slow, incremental, buggy, expensive rewrites of … literally everything?